DSP Robotics Support

Spogg wrote:In this scheme, a single “harmonic” would have a “bandwidth”, rather than being just a pure sinewave.

In some contexts that can be a feature rather than a bug.

In the late '80s/early '90s I worked on a system that used 2-op FM with low-rate random modulation to generate shaped background noise bands and "fat" tonals for a passive sonar simulator. By (rather carefully) selecting the statistics of the random modulators, you could get very fine shaping of the noise spectrum.

(The original system used an ASIC for the actual tone generation, with a single-board 68000 computer controlling the modulation. Later we ported that to a TMS320-based array processor.)

Next update, with freely configurable partial frequencies and a visual feedback. And a few more presets for inspiration.

Ok, I throw in the towel. I do not think anything else is going to be as good as this for some time. Martin this is fantastic and I can vision much more to come.The Cymbal and Gong are my favorites, but I am sure there will be many more presets and feature suggestions from us all still to come. Can't wait to see what the next rev brings!

Great stuff, BobF.....

I like the new version. Randomness and spreading helps making so much more sounds. But I don't like the very limited ways of editing a sound and the use of the inefficient original scope code. For example, I couldn't base any realtime drawing on this.
But that will probably get fixed over time! The synth itself is really good and produces awesome sounds (I had so much fun exploring gongs and metallic-like rythm sticks). As I said: I like it.

Amazing work Martin!
And wonderful sounds.

I listened carefully to the single sine and I noticed (on headphones) that there is a significant click approximately every 3 seconds, visible on a scope. This is not related to the note pitch or the detune amount. This is not a criticism, just hoping to help with feedback.

Also I feel I should mention that for some instruments, like bells for example, the fundamental can be at a lower level than F2, F3 and F4. In some cases there is literally no fundamental and our brains work it out and insert it by analysing the periodicity. So I hope we will get to see a manual editing feature so we can tailor the spectrum. It doesn’t have to be graphical; just enter a harmonic number and set the level, for fine tuning. As Hugh pointed out, it’s the first few harmonics that really define what we identify so I think that would be a really useful feature.

GO MARTIN!!

Cheers

Spogg

Thanks guys! I fixed the bug reported by Spogg and I also made the dispay a bit more efficient. Since the display does not need realtime updating, I hadn't geiven much thought to it.

@Tulamide: yes, the partials editing could be extended perhaps along the lines you suggested earlier in this thread. I can also think of replacing the envelopes with more flexible ones, add more filtering options, and maybe add an LFO.

When I try to go through the qwerty keyboard and I press Q,W,E,R,T,Y then U, I sounds like it switched two octaves higher?

Yeah, I'm getting the octaves in the wrong order from MIDI keyboard, the last two versions were like this.

I've been assuming it's just because I habitually use FS_64, and haven't actually tried it on 32-bit at all Maik's made some major mods to MIDI>Poly lately.

So I haven't reported it .. fugured that might make me a nuisance caller

H

Good catch, guys! The problem occurs with FS64 but not with FS3.0.8.1. I have tracked it down: the "Wave Write" prims (both Mono and Stereo) behave different in the FS versions. Will report to MyCo on Slack.

Edit: Done and fixed (PadOsc722 will play in both FS3.0.x and FS64 ).

Oops, you are right, my flowstone automatically opens it in the alpha, so now I specifically tried it in the old 3.0.6 and it works fine there, i guess thats why it wasnt reported by the others.